1. Field of the Invention
The present invention relates to an information processing technique of managing job information of a job to be executed by an external device.
2. Description of the Related Art
Recently, a business which opens a service to the public as a cloud service on the Internet has been developed. A print server is also required to provide a function as a print service on the Internet, that is, a cloud print service. This service releases a user from a large limitation that he or she cannot perform printing without connecting a terminal such as a personal computer (PC) to a printer and installing a specific printer driver to the terminal. In addition, the service provider can readily add server resources and improve design specifications in accordance with the load, that is, there are various merits for development.
A method using a database (to be referred to as a DB hereinafter) is available as a method of implementing a print job queue in a cloud print service like this. A print job having the highest priority in a desired printer can be decided by registering, in the DB, print jobs for different printers registered in the cloud print service together with information about, for example, the input date/time of each print job, and acquiring only a print job meeting appropriate conditions by using a database manipulation language such as SQL. Thus, print job queues of very many printers can be implemented by a single DB. This method has the merit that even an inexpensive printer incapable of implementing a print job queue by the hardware resources of the main body can implement a print job queue. In addition, many techniques for load balancing such as replication have been proposed for the DB. Since this can relatively easily enlarge the service scale, replication is one print job queue implementing method in the cloud print service.
Unidirectional replication is a technique for balancing the load on the DB. The unidirectional replication has an arrangement including a master DB for accepting a data update process, and one or more slave DBs for mirroring the state of the master DB. When compared to a DB configuration including only one DB, this arrangement can perform backup when a trouble occurs, and, in addition to that, can balance the access load on the master DB by performing access of reference only without any data update on the slave DB. The unidirectional replication is further classified into synchronous and asynchronous unidirectional replications in accordance with the data update process for the master DB, and the mirroring completion timing of the slave DB.
The synchronous replication ensures mirroring completion of the slave DB when data update to the master DB is complete, but has the demerit that the response speed of the data update process for the master DB decreases. In the asynchronous replication, the data update process for the master DB and mirroring for the slave DB are asynchronous, so the response speed of the data update process for the master DB is high. On the other hand, since a delay occurs in the completion of mirroring for the slave DB, a time during which data just registered in the master DB cannot be acquired from the slave DB is produced. Japanese Patent Laid-Open No. 2008-210164 has disclosed a technique by which a job is backed up in preparation for down of the DB, and, even when no data can be acquired from the DB, the job is executed without any interruption.
The method disclosed in Japanese Patent Laid-Open No. 2008-210164 is based on a recovery process when the DB is down. However, a job stored as backup data is not referred to unless the DB is down. In addition, the processing overhead is large because both a job transmission process and the backup process are performed.